118 results about "Aggregate function" patented technology

The invention relates to the technical field of artificial intelligence and databases, in particular to an NL2SQL method and system based on reinforcement learning, and the method comprises the following steps: natural language input preprocessing, semantic analysis and abstract representation construction, mode linking and candidate set determination, SQL generation, query execution and feedback optimization, and multi-round interaction processing. The method has the beneficial effects that complex scenes such as nested sub-query, multi-table JOIN and aggregation function combination are supported, and the SQL generation accuracy is improved by more than 30% compared with that of a traditional method (test data is from a spider data set); a mode link mechanism based on reinforcement learning can quickly adapt to a new database mode, and in a cross-domain test (such as switching from an e-commerce database to a medical database), the accuracy rate decrease amplitude is less than 15%.

The invention provides an intelligent number asking method for large model index query, and belongs to the field of natural language processing and database query. An auxiliary information base is constructed to store a database structure and historical question and answer examples, and key entities and fields are extracted by conducting structured format processing on natural language questions; a large language model guided by a small number of examples is utilized to generate an intermediate expression close to an SQL grammar and a plurality of SQL query candidates, and a consistency alignment mechanism is designed to verify candidate SQL in the aspects of field matching, aggregation functions, data tables and the like so as to screen out an optimal SQL query. Finally, the generated correct SQL and the corresponding question and answer pairs are fed back to the auxiliary information base, and self-learning enhancement is achieved. According to the method, under the condition that only a small number of training examples are needed, the accuracy and expandability of large language model text generation SQL query are improved.

A database query optimization method, comprising: acquiring a query statement, the query statement comprising a connection operation and a group aggregation operation; determining that a connection key and a group key in the query statement are the same, wherein an aggregation function in the query statement can be split, and an input parameter of the aggregation function comes from the same side as a connection condition in the query statement; creating a fusion operator, the fusion operator being used for completing the connection operation and the group aggregation operation; when a cost for executing the fusion operator is less than or equal to a target cost, on the basis of the fusion operator, generating an optimal execution plan related to the fusion operator, wherein the target cost is a cost for executing a connection operator and a group aggregation operator; and performing the operations step by step on the basis of an operation sequence in the optimal execution plan, so as to obtain a query result set. Therefore, when the query statement meets corresponding features, the connection operation and the group aggregation operation can be fused into one operator, thereby reducing the processing flow of an intermediate result, accelerating the execution of the query statement, and improving the query execution performance.

The database query optimization method comprises the steps that a query statement is obtained, and the query statement comprises a connection operation and a grouping aggregation operation; determining that a connection key and a grouping key in the query statement are the same, an aggregation function in the query statement is detachable, and input parameters of the aggregation function are from the same side of a connection condition in the query statement; creating a fusion operator, wherein the fusion operator is used for completing a connection operation and a grouping aggregation operation; under the condition that the cost of executing the fusion operator is smaller than or equal to a target cost, an optimal execution plan related to the fusion operator is generated according to the fusion operator, and the target cost is the cost of executing the connection operator and the grouping aggregation operator; and executing operations step by step according to the operation sequence in the optimal execution plan to obtain a query result set. Therefore, when the query statement meets the corresponding characteristics, the connection and grouping aggregation operations can be fused into one operator, so that the processing flow of an intermediate result can be reduced, the execution of the query statement is accelerated, and the execution performance of the query is improved.

The invention discloses a private network dynamic access control method and system, and relates to the technical field of network security and access control. The method comprises the following steps: acquiring equipment behavior data and network flow data in a private network, performing feature construction, and generating equipment trust features and network behavior features; and carrying out multi-dimensional risk assessment model training based on the equipment trust features and the network behavior features, carrying out real-time risk assessment on access requests or entities in the private network through the trained multi-dimensional risk assessment model, and generating a real-time risk score through a weighted aggregation function. According to the invention, through the multi-modal feature fusion model based on an attention mechanism, deep association of static attributes and dynamic behavior features of equipment is realized, and a real-time risk score is generated in combination with a multi-dimensional risk assessment model and a weighted aggregation algorithm. The limitation that in traditional access control, the evaluation dimension is single, the static strategy lags behind, and dynamic threats cannot be reflected is effectively overcome, and the accuracy and interpretability of private network access risk perception are remarkably improved.

Data structures and methods are described to provide a unique extendable group identifier for efficient aggregation query processing using multiple grouping keys. A method comprises retrieving a database query comprising an aggregate function of a selected column from a database table, grouped by a plurality of columns. The method further comprises maintaining a plurality of hash tables that map column values to grouping keys and identify a maximum grouping key. The method further comprises updating a plurality of bitmasks that define per-column grouping key bit positions, capable of storing the maximum grouping key, within a combined index value. The method further comprises allocating a result array sized according to the plurality of bitmasks. The method further comprises using the plurality of bitmasks and the plurality of hash tables to determine the combined index value to apply the aggregate function on the selected column for each row in the database table.

The invention belongs to the field of power system cascade fault prediction, and relates to a graph neural network power grid cascade fault prediction method based on physical information guidance, and the method comprises the steps: obtaining the operation data of a cascade fault, inputting the operation data of the cascade fault into a trained power grid cascade fault prediction model, and obtaining a prediction result; the power grid cascade fault prediction model comprises a common embedded graph neural network and a structure coding module; according to the invention, data enhancement is carried out by using a multi-step random walk matrix considering power grid line attributes and structure coding represented by node degrees, so that the capability of capturing topological information is improved; according to the invention, the interaction between the power grid nodes and the power grid lines is promoted by using the co-embedded graph neural network, and the accuracy of power grid cascade fault prediction is improved; according to the method, the node-edge interaction aggregation function of the co-embedded graph neural network is changed through the power transmission distribution factor of the power grid, the description capability of the actual complex interaction relation of the power grid is improved, and the accuracy of power grid cascade fault prediction is improved.

In one embodiment, a device identifies a set of attributes from telemetry data generated by one or more agents regarding an online application accessible via a network. The device provides an interactive display to a user interface that includes options for a user to specify a selection of one or more attributes from the set of attributes and to specify an aggregation function. The device updates the interactive display to show a visualization of the aggregation function applied to the selection of one or more attributes and configures the one or more agents to collect only a subset of the telemetry data based on the selection of the one or more attributes and the aggregation function.

Automatic generation of optimized aggregation metrics for usage-based insurance (UBI) includes performing a high-dimensional Bayesian optimization on a data archive, the data archive including a plurality of signals from vehicles and corresponding UBI effects, wherein the high-dimensional Bayesian optimization includes performing testing on weighted groups of the plurality of signals using one or more aggregation functions; and transmitting the one or more aggregation functions to a vehicle to cause the vehicle to provide aggregated signals for input to a UBI model to predict a UBI rate for the vehicle.

The invention relates to an APT detection method and system based on a multi-branch graph neural network, and the method comprises the steps: firstly constructing a traceability graph through analyzing a system log; secondly, performing semantic embedding on nodes in the traceability graph by adopting Word2Vec, and capturing context semantics of the nodes; through an embedding generation stage, semantic information of nodes is input into the trained multi-branch network model MBGNN to detect APT; the multi-branch network model comprises a GAT (GraphSAGE), a GraphSAGE and a GIN; the GAT branch uses an attention mechanism to model importance differences between nodes and neighbors; the GIN branch improves the discrimination ability of a local connection mode through a structure-sensitive aggregation function; and the GraphSAGE branch realizes feature aggregation through neighbor sampling. According to the method, efficient learning of global information of a large-scale traceability graph is realized, and the comprehensive recognition capability of APT behaviors and the adaptability to complex scenes are improved.

The invention discloses a federated backdoor defense method based on decoupling contrast learning, and the method comprises the steps: training a backdoor model based on a backdoor sample, and immediately stopping training after the backdoor model converges on the backdoor sample; respectively extracting a penultimate layer vector of the backdoor model and the local model from a sample pair held by the malicious client as a backdoor feature and a clean feature; comparing and learning the separated back door features and the clean features, and learning the clean features for the local model by using a sample weighting strategy to train the local model to obtain a trained local model; and sending local model parameters of the trained local model to a global server, and generating model parameters of a new global model based on the local model parameters through an aggregation function. The method aims at reducing information dependence between backdoor features and clean features through comparative learning, so that local model learning is free of backdoor representation, and the robustness of a global model is improved.

A graph processing engine is provided for executing a graph query comprising a parent query and a subquery nested within the parent query. The subquery uses a reference to one or more correlated variables from the parent query. Executing the graph query comprises initiating execution of the parent query, pausing the execution of the parent query responsive to the parent query matching the one or more correlated variables in an intermediate result set, generating a subquery identifier for each match of the one or more correlated variables, modifying the subquery to include a subquery aggregate function and a clause to group results by subquery identifier, executing the modified subquery using the intermediate result set and collecting subquery results into a subquery results table responsive to pausing execution of the parent query, and resuming execution of the parent query using the subquery results table.

According to the software supply chain dependency robustness evaluation method and device, the medium and the equipment provided by the invention, when robustness evaluation is performed on the open source software supply chain, the initial dependency graph is constructed according to the software dependency data from the plurality of data sources in the open source software supply chain, so that the space-time complexity of the dependency chain can be comprehensively restored; noise reduction is carried out on the initial dependency graph to form a noise reduction dependency graph by using a noise reduction auto-encoder and a graph neural network, and then multi-scale analysis is carried out to form an analysis result; the loss function of the noise reduction auto-encoder is formed by combining reconstruction loss and structure maintenance loss, and the graph neural network is trained by adopting a weighted aggregation function and a multi-task loss function and is evaluated through a structure disturbance experiment. Therefore, the robustness evaluation result obtained by performing multi-dimensional index evaluation on the analysis result can fully consider the influence of data noise, dynamic structure change and attack scenes, has relatively high accuracy, and provides powerful support for software supply chain security management.

The invention discloses a dynamic user behavior path analysis method, system and device and a medium. The method specifically comprises the following steps: receiving original behavior data through a registered custom aggregation function; inputting the original behavior data into a probability transfer model to generate a probability weighted path sequence; based on the probability weighted path sequence, mapping the path possibility ranking into a state space, and constructing a state-reward mapping relation by taking the user conversion rate as a reward signal; based on the state-reward mapping relation, a recommendation strategy is dynamically generated through a Q-learning algorithm, and the recommendation strategy is fed back to the service system in real time to be executed; and analyzing a path rule configured by a user in the business system through a Drools rule engine, executing hot update of the path rule, and feeding back the path rule after hot update to the probability transfer model. According to the method, dynamic, comprehensive, real-time and efficient user behavior path analysis is realized, a recommendation strategy can be accurately generated, and the method flexibly adapts to business rule changes.

The invention discloses a method and a system for solving cascade workshop multi-scene robust scheduling under a multi-objective framework, and the method comprises the following steps: constructing a mixed integer linear programming model considering the uncertainty of processing time, a robust scheduling problem in multiple scenes is converted into a multi-target problem taking the maximum completion time in each typical disturbance scene as an optimization target; designing a scene-driven heuristic construction method to generate a high-quality initial solution set; iterative search of a scheduling solution is carried out under an evolutionary greedy optimization framework based on scene decomposition, adaptive parameter control, subgroup division, local reconstruction and a variation guide strategy are fused, and the search capability and adaptability of an algorithm are improved; and finally, screening a Pareto optimal solution through an aggregation function and a niche mechanism. According to the method, multi-scene scheduling performance balance optimization is realized in a PCB manufacturing cascade production system, the method has the advantages of high modeling precision, high search efficiency, excellent solution set quality and the like, and the robustness and practicability of a scheduling scheme under a multi-scene disturbance condition are remarkably improved.

One or more embodiments of the invention provide a query statement optimization method and device, equipment and a storage medium. The method comprises the following steps: identifying each aggregation function in grouped query statements; for each aggregation function, judging whether parameters of the aggregation function can be calculated by a grouping expression or not; if yes, obtaining the number of parameter values of the aggregation function in each group; and under the condition that only one parameter value exists, representing the aggregation function according to parameters in the aggregation function to obtain an optimized query statement. According to the method, redundant grouping aggregation functions are eliminated, a better execution plan is obtained on the premise of ensuring the query equivalence, the overhead of the query request is reduced, and the query execution efficiency is improved.

The invention relates to a recommendation system in the field of artificial intelligence, and discloses a quaternion-based graph neural network social recommendation method. The method mainly comprises four parts: quaternion embedding, user preference acquisition, user explicit and implicit social relation fusion and score prediction. The method comprises the following steps: firstly, expressing user information, article information and user-article interaction information in the same space through quaternion embedding, and integrating the learned user and article information by utilizing the asymmetry property of a Hamilton product in a quaternion space and a weight sharing mechanism; and secondly, rotating user embedding and rotating article embedding are calculated in a quaternion space by utilizing non-exchangeability of a Hamilton product, more accurate user preference representation is obtained through an attention mechanism, and user social relations are distinguished. Secondly, modeling the relation between different objects in the social relation by connecting the corresponding real part and imaginary part in the quaternion; and fusing the explicit and implicit social relationships of the user by using the aggregation function. And finally, combining the user preference with the user social relationship, performing score prediction, and generating a recommendation list for the user. According to the method, the problem of high parameterization during recommendation model training can be effectively relieved, the representation ability of the model is enriched, and the accuracy of a recommendation result is improved.

The present invention discloses a multimodal travel mode fusion recommendation method based on a dynamic traffic network. The present invention first constructs the user's spatiotemporal travel trajectory into a heterogeneous information network, and considers the dynamic traffic network that is constantly changing during the travel process; pre-processes historical trajectory data according to the heterogeneous information network; extracts key meta-paths from the network mode corresponding to the heterogeneous information network to enrich the feature embedding of the object; realizes meta-path feature aggregation through aggregation functions; uses a meta-path-guided graph neural network to dynamically model users and travel modes, and obtains feature embedding with rich interactive information; the final feature embedding of users and travel modes is input into the MLP, fully learning the user's preference for travel modes in the spatiotemporal travel trajectory, and recommending travel modes that meet personalized needs to users. The present invention describes the user's personalized preference for travel modes in a fine-grained manner, while alleviating traffic pressure and improving traffic resilience.

The invention relates to the technical field of warehouse logistics management and artificial intelligence, in particular to a warehouse logistics dynamic scheduling optimization method and system based on deep learning, and the method comprises the steps: collecting static map data and dynamic sensor data of a warehouse environment; constructing a dynamic space-time diagram state at the current moment; in response to a business logic rule of the warehouse management system, triggering a scheduling optimization process, which comprises the following steps: generating a preliminary scheduling intention set by a constraint reinforcement learning scheduler; predicting to obtain a future space-time congestion thermodynamic diagram through a space-time diagram attention network predictor; quantitative processing is carried out through a self-defined aggregation function, and the future congestion cost of the scalar is determined; correcting the preliminary scheduling intention set by constraining a Lagrange solving framework of a reinforcement learning scheduler, and generating a final scheduling intention; outputting the final scheduling intention to a warehousing control system for execution; according to the method, the problem that decision making and prediction are disjointed in traditional scheduling is solved, and conversion from lagging response to active avoidance is achieved.

The invention relates to the technical field of fault diagnosis, in particular to a nuclear power station fault diagnosis method based on a multi-gating memory graph convolutional network, and the method comprises the steps: obtaining and processing fault data; embedding the gated memory unit into the graph convolutional network to construct a gated memory graph convolutional network; then, constructing a corresponding fault transfer graph according to the influence range of each fault, inputting fault data into each gated memory graph convolutional network embedded by the specific fault transfer graph in parallel to perform fault transfer feature extraction, and introducing memory and forgetting control for feature processing of each node in a time dimension through an embedded gating mechanism; extracting time and space features; and connecting and fusing the time and space features extracted from each sub-graph by using an attention mechanism aggregation function to generate final fault vector representation. Aiming at similar fault diagnosis of the same type, the model provided by the invention can effectively distinguish similar faults and obtain a good diagnosis result under the condition that other models cannot well distinguish the similar faults.

The invention relates to a data processing method and device, equipment, a readable storage medium and a program product. The method comprises the following steps: acquiring an operator and an aggregation function input by a user based on a preset programming interface; performing construction processing on a first input array and a second input array based on the operators to obtain an array matching relation table; wherein the first input array and the second input array are determined by performing abstract representation on protected private input data; and performing aggregation processing on elements in the array matching relation table based on the aggregation function, and determining a target aggregation result corresponding to the private input data. By adopting the method, the data privacy protection effect can be achieved.

The invention provides an SQL (Structured Query Language) generation method, system and equipment based on word segmentation and a knowledge graph and a medium, and the method comprises the following steps: judging the type of an input question, preprocessing the input question according to the type of the input question, and generating a standard input question; key information of the standard input question is extracted through a word segmentation labeling method, and standardization processing is conducted on the key information; an SQL statement is generated based on the key information by adopting a knowledge graph, and operation is performed according to the SQL statement to generate result information; performing customization processing and display on the generated result information; wherein the key information comprises time data, metadata, main data, an attribute value limiting condition, a data return condition, a sorting word and an aggregation function condition. According to the technical scheme provided by the invention, the natural language can be converted into the SQL language.

The present application relates to the technical field of database, provide a kind of Hash aggregation method and device based on Limit statement in openGuass, the method of the present application, comprising: according to query SQL statement generation Limit operator and HashAgg operator, wherein, Limit operator includes limit value, HashAgg operator includes aggregation function and grouping field information;Limit value in Limit operator is passed and saved to HashAgg operator;Execute query SQL, calculate the hash value of grouping field in data line, and construct HashBucket hash bucket according to the non-repeated number of occurrence of the hash value of grouping field and limit value;The aggregated field in the data of the same grouping field with the same hash value is placed in the same HashBucket hash bucket corresponding;After data scanning is completed, the aggregated field in each HashBucket hash bucket is aggregated by aggregation function, and the aggregation result is output to client.The present application can reduce the grouping and aggregation time of data in database and resource consumption as a whole.

The invention discloses a lane line smoothing processing method, device and equipment, a medium and a program product. The method comprises the following steps: determining a fitting position of a position point of a to-be-processed lane line, wherein the fitting position is represented by a plurality of fitting parameters; according to the fitting position, determining a shape loss function between a fitting line represented by the fitting position of the position point and the to-be-processed lane line; according to the position offset between the fitting position of the position point and the original position of the position point in the to-be-processed lane line, determining a flexible constraint condition corresponding to the position point; determining a parameter aggregation function of the flexible constraint parameter corresponding to the to-be-processed lane line based on the flexible constraint parameter corresponding to the position point of the to-be-processed lane line; and based on the flexible constraint condition, solving by combining the shape loss function and the parameter aggregation function to obtain the parameter value of the fitting parameter so as to obtain the smoothed lane line represented by the parameter value. According to the invention, the accuracy of smoothing the lane line can be improved.

The embodiment of the invention provides a power load prediction method and device based on a structure perception aggregation function, and the method comprises the steps: obtaining the node features of each node and the node features of a neighbor node in a power system topology model, aggregating the node features of each node and the node features of neighbor nodes in the power system topology model based on a structure sensing aggregation function to obtain the node features of each node after aggregation; acquiring historical time sequence features of each node in the power system topology model, and performing fusion processing on the historical time sequence features of each node and the aggregated node features to obtain fusion feature data of each node; and inputting the fusion feature data of each node into a trained deep neural network prediction model to obtain a power load prediction result of each node.

The present invention provides a method and system for optimizing reverse indexing in a relational database. The method reduces unnecessary comparison operations in intersections by using an inverted index posting list, uses a skip merge join method to skip rows through an aggregation function provided by the database, and then uses a multi-way join method to improve join performance and shorten the query process by simplifying complex join methods, so as to obtain the result of the posting list of keywords to be searched in one process, and at the same time process the joins of multiple posting lists to be searched, without obtaining the intermediate result value of the posting list intersection, thereby shortening the search time and improving the processing efficiency.

The application relates to the technical field of software test automation, and provides a requirement-driven test case automatic generation method and system based on a semantic test graph, which comprises the following steps: performing multidimensional static analysis on source code, respectively constructing an abstract syntax tree, a control flow graph and a function call graph, and extracting function call relations and data flow dependency relations; constructing a semantic test graph by taking functions as nodes, taking function call relations as main edges and taking data flow dependency relations as auxiliary edges, and aggregating function nodes in the semantic test graph into several function modules; obtaining several test requirement nodes, calculating semantic similarity between the test requirement nodes and the function modules, establishing a mapping relationship from the test requirement nodes to the function modules, and forming a requirement-program structure association graph; and based on the requirement-program structure association graph, generating a test intention for each test requirement node, and converting the test intention into an executable test case. The highly corresponding relationship between the generated test case and the business requirement is ensured.

The invention relates to the technical field of deep learning, and particularly discloses a neural network solving method and system for a weighted maximum satisfiability problem, and the method comprises the steps: representing an input conjunctive normal form formula as an edge splitting factor graph; constructing a graph neural network model, wherein the model comprises a supervised message passing layer and an unsupervised solution enhancement layer; an edge splitting factor graph is input to the graph neural network model, in the message passing layer, message passing is conducted on four types of edges through independent aggregation functions, and initial assignment prediction of variables is generated; and inputting the initial assignment prediction into the unsupervised solution enhancement layer, evaluating the quality of the current assignment, and carrying out iterative optimization on the variable assignment based on a relaxation optimization technology so as to output an optimized final solution. The invention provides a new graph representation method for a CNF formula, namely an edge splitting factor graph. According to the method, the spanning tree is generated, richer structure information is provided for the GNN, and therefore the learning efficiency is improved in the message passing process.

The invention provides a database query statement processing method and a related product. The method comprises the following steps: acquiring a database query statement; judging whether the database query statement has a deduplication aggregation function or not, wherein the deduplication aggregation function is an aggregation function with a deduplication property; if yes, counting the number of deduplication aggregation functions; and performing corresponding conversion operation on the database query statement according to the number of the deduplication aggregation functions, thereby eliminating the deduplication properties in the deduplication aggregation functions. By means of the method, after the aggregation function with the deduplication property is converted, the database can also calculate the aggregation function through the Hash grouping method, so that the situation that due to the fact that deduplication keywords exist in the aggregation function in an original statement, an optimizer cannot operate in parallel is avoided, and the processing performance of the database on the deduplication aggregation function is improved.

The embodiment of the application provides a kind of power load prediction method and device based on structure perception aggregation function, its method includes: obtaining the node feature of each node in the power system topology model and the node feature of its neighbor node, and based on structure perception aggregation function, the node feature of each node in the power system topology model and the node feature of its neighbor node are aggregated, to obtain the node feature after aggregation of each node;Obtain the historical time series feature of each node in the power system topology model, and by fusing the historical time series feature of each node with the aggregated node feature, obtain the fusion feature data of each node;By inputting the fusion feature data of each node into the trained deep neural network prediction model, the power load prediction result of each node is obtained.